Description
This type is an enumeration.
Usages
Allowed Values
| Value | Description |
|---|---|
ALL_MINILM_L6_V2 | The default "all-minilm-l6-v2" in-process embedding model |
ALL_MINILM_L6_V2_Q | The quantized variant of the "all-minilm-l6-v2" in-process embedding model, typically offering reduced memory footprint and potentially faster inference at some quality cost |
CUSTOM | A custom, user-provided in-process ONNX embedding model, when selected `path-to-model` and `path-to-tokenizer` needs to be provided |
See the manifest for all available types.